← back

Overview

A complete implementation of the Proximal Policy Optimization (PPO) algorithm from scratch in PyTorch, compatible with the OpenAI Gym / Gymnasium API.

Features

• Gymnasium-compatible. Works with any Gymnasium environment, discrete or continuous.
• Training utilities. Training scripts with progress tracking and visualization.
• Model persistence. Save and load trained models with automatic compatibility checks.

Technology

• RL framework: PyTorch, Gymnasium